Graphia is a powerful open-source visual analytics platform designed and developed to aid the interpretation of large and complex datasets.
Graphia can create and visualise networks from any table of numeric or discrete data values and display the often large and complex graph structures that result. It can also be used to visualise and analyse any data that is already in the form of a network.
- Many input data formats
- .csv, .tsv, .xls, .gml, .graphml, .json, .mat, .graphia
- Rapid calculation of correlation networks from primary data based on +ve and/or –ve correlations
- Pearson, Spearman rank
- Visualization of millions of data points and relationships
- Interactive visualizations in 2D and 3D
- Innovative node/edge attribute handling
- Fast, tunable network clustering
- MCL, Louvain
- Built in network analytics
- PageRank, Betweeness, Eccentricity, Enrichment Analysis
- Advanced network editing capabilities (filter on any attribute of nodes and edges)
- Integration of data from different platforms and analyses
- Customisable and simple to use web search
- Full range of options for network visualization, inspection and querying
- Easy export and sharing of analysis results
- Customizable to suit your data type or analysis needs
Graphia works with two broad classes of data; Network data where connections between entities are already known, and Numerical data where Graphia will construct a correlation graph of any table of numerical values. As an open-source project, if it doesn’t already do what you want it to, you can adapt it to your own needs.
When data is visualised in an intuitive and interactive manner, it allows the analyst the tackle certain problems whose size and complexity make them otherwise intractable. Graphia couples advanced computational algorithms with a visualisation interface that makes full use of the cognitive abilities of humans, providing deeper understanding and better communication of data.
Graphia is designed to be usable by any analyst on any computational platform. Its intuitive graphical user interface means that powerful analytical algorithms and searches are just a click away!
The world is full of large and complex data and Graphia has been designed to work with it. Every aspect of the tool has been optimised for maximum speed and performance, so you don’t have to wait.
Many analyses of complex data come down to a need to identify patterns and trends of interest. Graphia is designed to make this easy and quick by combining fast dynamic visualisations of data structure, with cluster analysis and the ability to explore the underlying data associated with the clusters observed.
Graphia provides a platform where different data types can be merged and explored as a graph, where connections or data attributes are derived from more than one source. A plugin architecture enables the user interface may be modified to suit a given data type, connect with predefined data sources or optimised for a given analytical process, allowing a user to tailor it to suit its specific analytical needs.
- Biological sciences – protein interaction data, transcriptomics, single cell analyses, proteomics, metabolomics, multiparameter flow cytometry, genotyping data, medical imaging data,
- Agri-tech – data relating to the performance of animals, crops, farms, etc.
- Fintech – any numerical data relating to changing variables over time, e.g. share prices or categorical data relating to the attributes of commercial entities
- Social media – network connections between individuals, companies, etc.
- Text mining – count matrices of words found across many documents
- Questionnaire – answers to questions are categorical (yes/no) or continuous (1-10)
Graphia is designed and built by a small dedicated group of scientists working in Edinburgh, Scotland. We are passionate about graphs, the power of visualisation and creating tools that aid the interpretation of complex data.
Our journey started 20 years ago, as we began to envisage how graphs could help us analyse the relationships between genes and proteins. For 15 years we developed our first software, BioLayout, a platform that taught us the power of graph-based analyses. This is still a great tool, but innate limitations in its design prevented us from developing it further, so in 2015 we set up a company, Kajeka Ltd, and started again. Unfortunately, the company did not survive but Graphia, the main product of Kajeka did and we have now made it free and open source for all to enjoy.
To reference Graphia, please cite:….
There are lots of new features and functionality we still wish to add to Graphia and envisage a desire from some to integrate it into local data infrastructures or web resources. If you are interested to work with us to further improve Graphia, or need our help in other ways, we would love to hear from you.
Please contact: [email protected]
If you use Graphia and would like to support the project we really need funding, as at the moment we have none. No money = no support or development.
Please help us to help you!
We would like to thank all those who have helped us on our journey: previous members of the team - Seb Horsewell; investors in Kajeka – Les Gaw, David Hume, Paul Chowdhry, Paul Gregory, Old College Capital, Scottish Investment Bank; funders – Scottish Enterprise, and finally, The Roslin Institute, University of Edinburgh and the BBSRC.
